1 /* GNU tar Archive Format description.
2
3 Copyright 1988-2023 Free Software Foundation, Inc.
4
5 This file is part of GNU tar.
6
7 GNU tar is free software; you can redistribute it and/or modify
8 it under the terms of the GNU General Public License as published by
9 the Free Software Foundation; either version 3 of the License, or
10 (at your option) any later version.
11
12 GNU tar is distributed in the hope that it will be useful,
13 but WITHOUT ANY WARRANTY; without even the implied warranty of
14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 GNU General Public License for more details.
16
17 You should have received a copy of the GNU General Public License
18 along with this program. If not, see <http://www.gnu.org/licenses/>. */
19
20 /* tar Header Block, from POSIX 1003.1-1990. */
21
22 /* POSIX header. */
23
24 struct posix_header
25 { /* byte offset */
26 char name[100]; /* 0 */
27 char mode[8]; /* 100 */
28 char uid[8]; /* 108 */
29 char gid[8]; /* 116 */
30 char size[12]; /* 124 */
31 char mtime[12]; /* 136 */
32 char chksum[8]; /* 148 */
33 char typeflag; /* 156 */
34 char linkname[100]; /* 157 */
35 char magic[6]; /* 257 */
36 char version[2]; /* 263 */
37 char uname[32]; /* 265 */
38 char gname[32]; /* 297 */
39 char devmajor[8]; /* 329 */
40 char devminor[8]; /* 337 */
41 char prefix[155]; /* 345 */
42 /* 500 */
43 };
44
45 #define TMAGIC "ustar" /* ustar and a null */
46 #define TMAGLEN 6
47 #define TVERSION "00" /* 00 and no null */
48 #define TVERSLEN 2
49
50 /* Values used in typeflag field. */
51 #define REGTYPE '0' /* regular file */
52 #define AREGTYPE '\0' /* regular file */
53 #define LNKTYPE '1' /* link */
54 #define SYMTYPE '2' /* reserved */
55 #define CHRTYPE '3' /* character special */
56 #define BLKTYPE '4' /* block special */
57 #define DIRTYPE '5' /* directory */
58 #define FIFOTYPE '6' /* FIFO special */
59 #define CONTTYPE '7' /* reserved */
60
61 #define XHDTYPE 'x' /* Extended header referring to the
62 next file in the archive */
63 #define XGLTYPE 'g' /* Global extended header */
64
65 /* Bits used in the mode field, values in octal. */
66 #define TSUID 04000 /* set UID on execution */
67 #define TSGID 02000 /* set GID on execution */
68 #define TSVTX 01000 /* reserved */
69 /* file permissions */
70 #define TUREAD 00400 /* read by owner */
71 #define TUWRITE 00200 /* write by owner */
72 #define TUEXEC 00100 /* execute/search by owner */
73 #define TGREAD 00040 /* read by group */
74 #define TGWRITE 00020 /* write by group */
75 #define TGEXEC 00010 /* execute/search by group */
76 #define TOREAD 00004 /* read by other */
77 #define TOWRITE 00002 /* write by other */
78 #define TOEXEC 00001 /* execute/search by other */
79
80 /* tar Header Block, GNU extensions. */
81
82 /* In GNU tar, SYMTYPE is for to symbolic links, and CONTTYPE is for
83 contiguous files, so maybe disobeying the "reserved" comment in POSIX
84 header description. I suspect these were meant to be used this way, and
85 should not have really been "reserved" in the published standards. */
86
87 /* *BEWARE* *BEWARE* *BEWARE* that the following information is still
88 boiling, and may change. Even if the OLDGNU format description should be
89 accurate, the so-called GNU format is not yet fully decided. It is
90 surely meant to use only extensions allowed by POSIX, but the sketch
91 below repeats some ugliness from the OLDGNU format, which should rather
92 go away. Sparse files should be saved in such a way that they do *not*
93 require two passes at archive creation time. Huge files get some POSIX
94 fields to overflow, alternate solutions have to be sought for this. */
95
96 /* Descriptor for a single file hole. */
97
98 struct sparse
99 { /* byte offset */
100 char offset[12]; /* 0 */
101 char numbytes[12]; /* 12 */
102 /* 24 */
103 };
104
105 /* Sparse files are not supported in POSIX ustar format. For sparse files
106 with a POSIX header, a GNU extra header is provided which holds overall
107 sparse information and a few sparse descriptors. When an old GNU header
108 replaces both the POSIX header and the GNU extra header, it holds some
109 sparse descriptors too. Whether POSIX or not, if more sparse descriptors
110 are still needed, they are put into as many successive sparse headers as
111 necessary. The following constants tell how many sparse descriptors fit
112 in each kind of header able to hold them. */
113
114 #define SPARSES_IN_EXTRA_HEADER 16
115 #define SPARSES_IN_OLDGNU_HEADER 4
116 #define SPARSES_IN_SPARSE_HEADER 21
117
118 /* Extension header for sparse files, used immediately after the GNU extra
119 header, and used only if all sparse information cannot fit into that
120 extra header. There might even be many such extension headers, one after
121 the other, until all sparse information has been recorded. */
122
123 struct sparse_header
124 { /* byte offset */
125 struct sparse sp[SPARSES_IN_SPARSE_HEADER];
126 /* 0 */
127 char isextended; /* 504 */
128 /* 505 */
129 };
130
131 /* The old GNU format header conflicts with POSIX format in such a way that
132 POSIX archives may fool old GNU tar's, and POSIX tar's might well be
133 fooled by old GNU tar archives. An old GNU format header uses the space
134 used by the prefix field in a POSIX header, and cumulates information
135 normally found in a GNU extra header. With an old GNU tar header, we
136 never see any POSIX header nor GNU extra header. Supplementary sparse
137 headers are allowed, however. */
138
139 struct oldgnu_header
140 { /* byte offset */
141 char unused_pad1[345]; /* 0 */
142 char atime[12]; /* 345 Incr. archive: atime of the file */
143 char ctime[12]; /* 357 Incr. archive: ctime of the file */
144 char offset[12]; /* 369 Multivolume archive: the offset of
145 the start of this volume */
146 char longnames[4]; /* 381 Not used */
147 char unused_pad2; /* 385 */
148 struct sparse sp[SPARSES_IN_OLDGNU_HEADER];
149 /* 386 */
150 char isextended; /* 482 Sparse file: Extension sparse header
151 follows */
152 char realsize[12]; /* 483 Sparse file: Real size*/
153 /* 495 */
154 };
155
156 /* OLDGNU_MAGIC uses both magic and version fields, which are contiguous.
157 Found in an archive, it indicates an old GNU header format, which will be
158 hopefully become obsolescent. With OLDGNU_MAGIC, uname and gname are
159 valid, though the header is not truly POSIX conforming. */
160 #define OLDGNU_MAGIC "ustar " /* 7 chars and a null */
161
162 /* The standards committee allows only capital A through capital Z for
163 user-defined expansion. Other letters in use include:
164
165 'A' Solaris Access Control List
166 'E' Solaris Extended Attribute File
167 'I' Inode only, as in 'star'
168 'N' Obsolete GNU tar, for file names that do not fit into the main header.
169 'X' POSIX 1003.1-2001 eXtended (VU version) */
170
171 /* This is a dir entry that contains the names of files that were in the
172 dir at the time the dump was made. */
173 #define GNUTYPE_DUMPDIR 'D'
174
175 /* Identifies the *next* file on the tape as having a long linkname. */
176 #define GNUTYPE_LONGLINK 'K'
177
178 /* Identifies the *next* file on the tape as having a long name. */
179 #define GNUTYPE_LONGNAME 'L'
180
181 /* This is the continuation of a file that began on another volume. */
182 #define GNUTYPE_MULTIVOL 'M'
183
184 /* This is for sparse files. */
185 #define GNUTYPE_SPARSE 'S'
186
187 /* This file is a tape/volume header. Ignore it on extraction. */
188 #define GNUTYPE_VOLHDR 'V'
189
190 /* Solaris extended header */
191 #define SOLARIS_XHDTYPE 'X'
192
193 /* J@"org Schilling star header */
194
195 struct star_header
196 { /* byte offset */
197 char name[100]; /* 0 */
198 char mode[8]; /* 100 */
199 char uid[8]; /* 108 */
200 char gid[8]; /* 116 */
201 char size[12]; /* 124 */
202 char mtime[12]; /* 136 */
203 char chksum[8]; /* 148 */
204 char typeflag; /* 156 */
205 char linkname[100]; /* 157 */
206 char magic[6]; /* 257 */
207 char version[2]; /* 263 */
208 char uname[32]; /* 265 */
209 char gname[32]; /* 297 */
210 char devmajor[8]; /* 329 */
211 char devminor[8]; /* 337 */
212 char prefix[131]; /* 345 */
213 char atime[12]; /* 476 */
214 char ctime[12]; /* 488 */
215 /* 500 */
216 };
217
218 #define SPARSES_IN_STAR_HEADER 4
219 #define SPARSES_IN_STAR_EXT_HEADER 21
220
221 struct star_in_header
222 {
223 char fill[345]; /* 0 Everything that is before t_prefix */
224 char prefix[1]; /* 345 t_name prefix */
225 char fill2; /* 346 */
226 char fill3[8]; /* 347 */
227 char isextended; /* 355 */
228 struct sparse sp[SPARSES_IN_STAR_HEADER]; /* 356 */
229 char realsize[12]; /* 452 Actual size of the file */
230 char offset[12]; /* 464 Offset of multivolume contents */
231 char atime[12]; /* 476 */
232 char ctime[12]; /* 488 */
233 char mfill[8]; /* 500 */
234 char xmagic[4]; /* 508 "tar" */
235 };
236
237 struct star_ext_header
238 {
239 struct sparse sp[SPARSES_IN_STAR_EXT_HEADER];
240 char isextended;
241 };
242
243 /* END */
244
245
246 /* tar Header Block, overall structure. */
247
248 /* tar files are made in basic blocks of this size. */
249 #define BLOCKSIZE 512
250
251 enum archive_format
252 {
253 DEFAULT_FORMAT, /* format to be decided later */
254 V7_FORMAT, /* old V7 tar format */
255 OLDGNU_FORMAT, /* GNU format as per before tar 1.12 */
256 USTAR_FORMAT, /* POSIX.1-1988 (ustar) format */
257 POSIX_FORMAT, /* POSIX.1-2001 format */
258 STAR_FORMAT, /* Star format defined in 1994 */
259 GNU_FORMAT /* Same as OLDGNU_FORMAT with one exception:
260 see FIXME note for to_chars() function
261 (create.c:189) */
262 };
263
264 /* Information about a sparse file. */
265 struct sp_array
266 {
267 off_t offset;
268 off_t numbytes;
269 };
270
271 struct xheader
272 {
273 struct obstack *stk;
274 size_t size;
275 char *buffer;
276 uintmax_t string_length;
277 };
278
279 /* Information about xattrs for a file. */
280 struct xattr_array
281 {
282 char *xkey;
283 char *xval_ptr;
284 size_t xval_len;
285 };
286
287 struct xattr_map
288 {
289 struct xattr_array *xm_map;
290 size_t xm_size; /* Size of the xattr map */
291 size_t xm_max; /* Max. number of entries in xattr_map */
292 };
293
294 struct tar_stat_info
295 {
296 char *orig_file_name; /* name of file read from the archive header */
297 char *file_name; /* name of file for the current archive entry
298 after being normalized. */
299 bool had_trailing_slash; /* true if the current archive entry had a
300 trailing slash before it was normalized. */
301 char *link_name; /* name of link for the current archive entry. */
302
303 char *uname; /* user name of owner */
304 char *gname; /* group name of owner */
305
306 char *cntx_name; /* SELinux context for the current archive entry. */
307
308 char *acls_a_ptr; /* Access ACLs for the current archive entry. */
309 size_t acls_a_len; /* Access ACLs for the current archive entry. */
310
311 char *acls_d_ptr; /* Default ACLs for the current archive entry. */
312 size_t acls_d_len; /* Default ACLs for the current archive entry. */
313
314 struct stat stat; /* regular filesystem stat */
315
316 /* STAT doesn't always have access, data modification, and status
317 change times in a convenient form, so store them separately. */
318 struct timespec atime;
319 struct timespec mtime;
320 struct timespec ctime;
321
322 off_t archive_file_size; /* Size of file as stored in the archive.
323 Equals stat.st_size for non-sparse files */
324
325 bool is_sparse; /* Is the file sparse */
326
327 /* For sparse files: */
328 unsigned sparse_major;
329 unsigned sparse_minor;
330 size_t sparse_map_avail; /* Index to the first unused element in
331 sparse_map array. Zero if the file is
332 not sparse */
333 size_t sparse_map_size; /* Size of the sparse map */
334 struct sp_array *sparse_map;
335
336 off_t real_size; /* The real size of sparse file */
337 bool real_size_set; /* True when GNU.sparse.realsize is set in
338 archived file */
339
340 bool sparse_name_done; /* Set to true if 'GNU.sparse.name' header was
341 processed pax header parsing. Following 'path'
342 header (lower priority) will be ignored. */
343
344 struct xattr_map xattr_map;
345
346 /* Extended headers */
347 struct xheader xhdr;
348
349 /* For dumpdirs */
350 bool is_dumpdir; /* Is the member a dumpdir? */
351 bool skipped; /* The member contents is already read
352 (for GNUTYPE_DUMPDIR) */
353 char *dumpdir; /* Contents of the dump directory */
354
355 /* Parent directory, if creating an archive. This is null if the
356 file is at the top level. */
357 struct tar_stat_info *parent;
358
359 /* Directory stream. If this is not null, it is in control of FD,
360 and should be closed instead of FD. */
361 DIR *dirstream;
362
363 /* File descriptor, if creating an archive, and if a directory or a
364 regular file or a contiguous file.
365
366 It is zero if no file descriptor is available, either because it
367 was never needed or because it was open and then closed to
368 conserve on file descriptors. (Standard input is never used
369 here, so zero cannot be a valid file descriptor.)
370
371 It is negative if it could not be reopened after it was closed.
372 Negate it to find out what errno was when the reopen failed. */
373 int fd;
374
375 /* Exclusion list */
376 struct exclist *exclude_list;
377 };
378
379 union block
380 {
381 char buffer[BLOCKSIZE];
382 struct posix_header header;
383 struct star_header star_header;
384 struct oldgnu_header oldgnu_header;
385 struct sparse_header sparse_header;
386 struct star_in_header star_in_header;
387 struct star_ext_header star_ext_header;
388 };